Recent Advances in Representation Learning for Electronic Health Records: A Systematic Review

Abstract

Representation Learning (RL) aims to convert data into low-dimensional and dense real-valued vectors, so as to realize reasoning in vector space. RL is one of the important research contents in the analysis of health data. This paper systematically reviews the latest research on Electronic Health Records (EHR) RL. We searched the Web of Science, Google Scholar, and Association for Computing Machinery Digital Library for papers involving EHR RL. On the basis of literature review, we propose a new taxonomy to categorize the state-of-the-art EHR RL methods into three categories: statistics learning-based RL methods, knowledge RL methods and graph RL methods. We analyze and summarize their characteristics according to the input data form and underlying learning mechanisms. In addition, we provide evaluation strategies to verify the quality of EHR representations from both intrinsic and extrinsic perspectives. Finally, we put forward three promising research directions to promote future research. Overall, this survey aims to provide a profound overview of state-of-the-art developments in the field of EHR RL and to help researchers find the most appropriate methods.

Analyzing wav2vec embedding in Parkinson’s disease speech: A study on cross-database classification and regression tasks

“…Recently, representation learning (RL) technology has facilitated the transformation of raw biomedical data into compact and low-dimensional vectors, commonly referred to as embeddings [34]. These embedding methods have gained widespread adoption across various biomedical domains, including natural language processing (NLP), where they are utilized to represent clinical concepts derived from unstructured clinical notes, such as symptoms and lab test results [35].…”

A Review on Electronic Health Record Text-Mining for Biomedical Name Entity Recognition in Healthcare Domain

“…Information extraction in biomedical informatics is essential for handling medical literature, which grows exponentially and extracts information in medical research [1]. Biomedical informatics analyze and identify biomedical name entity recognition (bNER) and healthcare as biomedical entities in electronic health records (EHRs), such as drugs (medicines), proteins from compound genes, and diseases in unstructured medical texts [2][3][4][5][6]. With the unprecedented expansion of electronic medical records (EMR) worldwide, millions of pieces of data have been collected since the publication of the basic norms of EMR [7][8][9][10].…”